154 Royal College of Physicians, 2011. All rights reserved.

I CONFERENCE SUMMARIES Clinical Medicine 2011, Vol 11, No 2: 1546

Introduction
Diabetic ketoacidosis (DKA) is the metabolic abnormality of
type 1 diabetes. Near patient testing for the ketones is now
readily available for monitoring allowing for a shift away from
using glucose levels to drive treatment decisions in the manage-
ment of DKA. National UK guidelines have been developed to
reflect the development in technology.
1
They are evidence based
where possible but are also drawn from accumulated multipro-
fessional knowledge and consensus agreement.
Pathophysiology and diagnosis
DKA usually occurs as a consequence of insulin deficiency. This
enhances hepatic glucose production. Enhanced fat breakdown
increases serum free fatty acids that are then metabolised pro-
ducing large quantities of ketones and metabolic acidosis. Fluid
depletion is a serious problem caused by osmotic diuresis.
Electrolyte shifts and depletion are, in part, related to the osmotic
diuresis. Hyper- and hypokalaemia are particular dangers.
Absolute diagnostic criteria do not exist, however, the fol-
lowing are proposed: ketonaemia 3 mmol/l by ketone meter,
or significant ketonuria (more than 2 on standard urine
sticks); blood glucose over 11 mmol/l or known diabetes mel-
litus; and venous bicarbonate (HCO
3
) 15 mmol/l and/or
venous pH 7.3.
Developments in management
Recent developments now allow us to focus on the underlying
metabolic abnormality (ketonaemia), which simplifies the treat-
ment of those who present with modest elevation of blood glu-
cose but with acidosis secondary to ketonaemia euglycaemic
diabetic ketoacidosis.
2,3
Blood glucose is routinely checked at the bedside, but portable
ketone meters now also allow bedside measurement of 3-beta-
hydroxybutyrate.
47
The resolution of DKA depends upon the
suppression of ketonaemia, therefore the measurement of blood
ketones now represents best practice in monitoring the response
to treatment.
8
Access to blood gas and blood electrolyte mea-
surements is now usually available within a few minutes of
blood being taken. Therefore glucose, ketones and electrolytes,
including bicarbonate and venous pH, should be assessed at, or
near, the bedside.
The involvement of diabetes specialist teams
Diabetes specialist team (DST) involvement shortens patient
stay and improves outcomes.
9
All patients should be reviewed by
a member of the DST prior to discharge, not undertaking this is
a governance issue.
10
General management issues
There is universal agreement that the most important initial
therapeutic intervention is appropriate fluid replacement fol-
lowed by insulin administration. The main aims of fluid replace-
ment are to:

restore circulatory volume

clear ketones

correct electrolyte imbalance.

In patients with kidney failure or heart failure, as well as the
elderly and adolescents, the rate and volume of fluid replace-
ment may need to be modified. The aim of the first few litres of
fluid is to correct any hypotension, replenish the intravascular
deficit, and counteract the effects of the osmotic diuresis with
rectification of electrolyte disturbance.
Assessment of severity
The presence of one or more of the following may indicate
severe DKA and admission to a level 2/high dependency unit
environment; insertion of a central line and immediate senior
review should be considered:

blood ketones over 6 mmol/l

bicarbonate level below 5 mmol/l

venous/arterial pH below 7.1

hypokalaemia on admission (under 3.5 mmol/l)

Glasgow Coma Scale less than 12 or abnormal AVPU (Alert,

Voice, Pain, Unresponsive) scale

oxygen saturation below 92% on air (assuming normal

baseline respiratory function)

systolic blood pressure below 90 mmHg

pulse over 100 or below 60 bpm.

Insulin therapy and metabolic treatment targets
A fixed rate intravenous insulin infusion (FRIVII) calculated
on 0.1 units/kg infusion is recommended. Patient demo-
graphics are changing and patients with DKA are now more
likely to be obese or suffering with other insulin-resistant
MW Savage, consultant physician in diabetes and endocrinology,
Diabetes and Endocrinology Centre, North Manchester Diabetes Centre,
Manchester
Management of diabetic ketoacidosis
MW Savage
Management of diabetic ketoacidosis
Royal College of Physicians, 2011. All rights reserved. 155
states, including pregnancy. This has led to the re-emergence of
FRIVII in adults in the USA and international paediatric prac-
tice.
1113
The FRIVII may need to be adjusted; the recom-
mended targets are:

reduction of the blood ketone concentration by 0.5

mmol/l/hour

if ketone meter not available, the venous bicarbonate should

rise by 3 mmol/l/hour and capillary blood glucose fall by
3 mmol/l/hour

potassium should be maintained between 4.0 and

5.0 mmol/l.
Intravenous glucose concentration
The management should be focused on clearing ketones as well
as normalising blood glucose. It is often necessary to administer
an intravenous (iv) infusion of 10% glucose via an iv pump in
order to avoid hypoglycaemia and permit the continuation of a
FRIVII to suppress ketogenesis. Introduction of 10% glucose is
recommended when the blood glucose falls below 14 mmol/l. It
is important to continue 0.9% sodium chloride solution concur-
rently via an iv pump.
Continuation of long-acting insulin analogues
In the last few years the use of long-acting basal insulin ana-
logues (insulin detemir Levemir, insulin glargine Lantus)
has become widespread. Continuation of long-acting analogues
during the initial management of DKA provides background
insulin when the iv insulin is discontinued. This avoids rebound
hyperglycaemia when iv insulin is stopped and should avoid
excess length of stay.
Serious complications of DKA and its treatment
Hypokalaemia and hyperkalaemia are potentially life-threat-
ening conditions during the management of DKA. There is a
risk of acute renal failure associated with severe dehydration and
it is therefore recommended that no potassium be prescribed with
the initial fluid resuscitation or if the serum potassium level
remains above 5.5 mmol/l. However, potassium will almost always
fall as the DKA is treated with insulin, thus it is recommended that
0.9% sodium chloride solution with potassium 40 mmol/l
(ready-mixed) is prescribed as long as the serum potassium level
is below 5.5 mmol/l and the patient is passing urine.
If the serum potassium level falls below 3.5 mmol/l the potas-
sium regimen needs review. All aspects of potassium use must
comply with local and national guidance.
14
Trusts need to
ensure that they have local protocols in place, which allow for
the safe administration of concentrated potassium injectables.
Other serious complications include cerebral and pulmonary
oedema which are discussed in the online document.
1
Close
attention to fluid balance and slower fluid replacement is recom-
mended for younger adults; for children 18 years of age use
the British Society for Paediatric Endocrinology and Diabetes
DKA guidelines which can be found at: www.bsped.org.uk/
professional/guidelines/docs/DKAGuideline.pdf
Hypoglycaemia
The blood glucose may fall very rapidly as ketoacidosis is cor-
rected and a common mistake is to allow it to drop to hypogly-
caemic levels. This may result in a rebound ketosis driven by
counter-regulatory hormones. Severe hypoglycaemia is also
associated with cardiac arrhythmias, acute brain injury
and death. It is partly for this reason that 10% glucose is
recommended.
Conclusion and summary
DKA is a medical emergency with a significant morbidity and
mortality. It is now recommended that FRIVII be used with bed-
side measurement of metabolic parameters. The DST should
always be involved as soon as possible and ideally within
24 hours because this has been demonstrated to be associated
with a better patient experience and reduced length of stay.
In the management of diabetic ketoacidosis the following
guidance should therefore be followed:

measure blood ketones, venous (not arterial) pH and bicar-

bonate and use results as treatment markers

monitor ketones and glucose using bedside meters, when

available and operating within their quality assurance range

monitor electrolytes on blood gas analysers with intermit-

tent laboratory confirmation

replace sliding scale insulin with weight-based FRIVII

involve the DST as soon as possible

continue long-acting insulin analogues as normal.

1
References
1 Joint British Diabetes Societies Inpatient Care Group. The manage-
ment of diabetic ketoacidosis in adults. London: NHS Diabetes, 2010.
www.diabetes.nhs.uk/document.php?o=1336
2 Munro JF, Campbell IW, Mccuish AC, Duncan LJP. Euglycaemic dia-
betic ketoacidosis. BMJ 1973;2:57880.
3 Jenkins D, Close CE, Krentz AJ, Nattrass M, Wright AD.
Euglycaemic diabetic ketoacidosis: does it exist? Acta Diabetol
1993;30:2513.
4 Sheikj-Ali M, Karon BS, Basu A et al. Can serum beta-hydroxybutyrate
be used to diagnose diabetic ketoacidosis? Diab Care 2008;4:6437.
5 Bektas F, Eray O, Sari R, Akbas H. Point of care testing of diabetic
patients in the emergency department. Endocr Res 2004;30:395402.
6 Khan ASA, Talbot JA, Tiezen KL et al. Evaluation of a bedside blood
ketone sensor: the effects of acidosis, hyperglycaemia and acetoacetate
on sensor performance. Diab Med 2004;21:7825.
7 Wallace TM, Matthews DR. Recent advances in the monitoring and
management of diabetic ketoacidosis. QJM2004;97:77380.
8 Wiggam MI, OKane MJ, Harper R et al. Treatment of diabetic
ketoacidosis using normalization of blood 3-hydroxybutyrate concen-
tration as the endpoint of emergency management. A randomized
controlled study. Diab Care 1997;20:134752.
MW Savage
156 Royal College of Physicians, 2011. All rights reserved.
9 Cavan DA, Hamilton P, Everett J, Kerr D. Reducing hospital inpatient
length of stay for patients with diabetes. Diab Med 2001;18:1624.
10 Clement S, Braithwaite SS, Magee MF et al. Management of diabetes
and hyperglycemia in hospitals. Diab Care 2004;27:55391.
11 Kitabchi, AE, Umpierrez, GE, Miles, JM, Fisher, JN. Hyperglycemic
crises in adult patients with diabetes: a consensus statement from the
American Diabetes Association. Diab Care 2009;32:1335.
12 British Society for Paediatric Endocrinology and Diabetes (BSPED)
guidelines for the management of DKA. www.bsped.org.uk/profes-
sional/guidelines/docs/DKAGuideline.pdf
13 International Society for Pediatric and Adolescent Diabetes. (IPSAD) 2007.
www.ispad.org/FileCenter/10-wolfsdorf_Ped_Diab_2007,8.28-43.pdf
14 National Patient Safety Authority. Patient Safety Alert. Potassium solu-
tions: risks to patients from errors occurring during intravenous adminis-
tration. London: NPSA, 2002.
Address for correspondence: Dr MW Savage,
Diabetes and Endocrinology Centre, North Manchester
Diabetes Centre, Manchester, M8 5RB.
Email: mark.savage@pat.nhs.uk
Acute pulmonary embolism (PE) is characterised by obstruction
of the pulmonary arterial tree by thrombi commonly origi-
nating from the leg or pelvic veins. Although deep vein throm-
bosis (DVT) and PE are manifestations of the same disease,
mortality is higher in those who present with PE.
1
In 2005, it was
estimated that 25,000 people in the UK die of preventable hos-
pital-acquired venous thromboembolism (VTE) every year.
Community studies have reported an annual incidence for VTE
of 1.41.8 per 1,000 with a third presenting with PE. The inci-
dence of VTE rises markedly with age; over the age of 75, the
annual incidence reached 1 per 100 population.
2
Ten per cent of acute PE are immediately fatal, with a further
5% mortality despite treatment. Perfusion defects resolve in
50% after one month of treatment, with complete resolution in
two-thirds.
3
There is a two-year cumulative incidence of 3.8%
for chronic thromboembolic pulmonary hypertension due to
poor resolution and/or a pulmonary arteriopathy.
4
Six to 15% of
patients have recurrent thrombosis after anticoagulation partic-
ularly in those with permanent risk factors.
Risk factors
The risk factors for development of VTE are venous stasis, vas-
cular damage and hypercoaguability. Predisposing factors
include reduced mobility, medical co-morbidities, surgery,
active cancer, previous VTE, drugs, pregnancy, central lines,
thrombophilias, obesity and varicose veins. In approximately
20% idiopathic PE may occur.
Symptoms and signs
The symptoms include dyspnoea, pleuritic or retrosternal chest
pain, cough, haemoptysis and syncope. Clinically the patient
may be tachypnoeic, tachycardic, and hypoxic with an elevated
jugular venous pressure, a gallop rhythm, a widely split second
heart sound, a tricuspid regurgitation murmur and possible
flow murmurs around the embolic obstructions. With pul-
monary infarction there may be a pleural rub and pyrexia. In
severe cases, the right ventricle fails to cope with the increase in
afterload resulting in systemic hypotension and shock. Signs of
a DVT increase the index of suspicion for PE.
Investigation of patients with suspected PE
The symptoms and signs in acute PE are neither sensitive nor
specific and only a third of referred patients will have PE. Hence
a pre-test assessment of clinical probability should be under-
taken in all cases. Several clinical predictive models based on risk
factors, symptoms and signs have been developed. The Wells
prediction model for PE has been validated and is the most
widely used.
5
As shown in Fig 1, PE can be excluded in patients
who have a low clinical probability and a negative D-dimer
without further investigations.
Investigations
Plasma D-dimer gives a measure of endogenous fibrinolysis. It
has a good negative predictive value and is useful in excluding
PE. It is not diagnostic as it has a low specificity and is elevated
in a wide range of conditions including infection, trauma and
cancer. It should not be measured in patients with a high clinical
Karen KK Sheares, respiratory consultant, Pulmonary Vascular Diseases
Unit, Papworth Hospital NHS Foundation Trust, Cambridge
How do I manage a patient with suspected acute pulmonary
embolism?
Karen KK Sheares
Clinical Medicine 2011, Vol 11, No 2: 1569